Seely and Duong (2001) start from the statistic from the Standish Group stating that nearly 90% of the sampled projects failed. From there they developed a framework in the form of The Dynamic Baseline Model (DBM). The DBM model has two facets, people and projects.
From a people standpoint, they define four levels of learning for a Project Manager , Management by Rules, by Methods, by Objectives, or by Values. They even give a %ge of the PM population that could fit in each level including character types and traits.
On the project front, the Lowest Management Level is the key control point for projects and where the projects must be managed to deal effectively with issues that may arise. For a production project the LML is “supervisor”, construction a “manager”, development is a “director” and “evolution” an owner.
The main issues here could be the fact that it only take two angles but does not offer an holistic perspective of root causes of failures. Kerzner (2009) defines that “Perceived failure is the net sum of actual failure and planning failure”, he considers that bottom line these failures are mainly due to lack of risk management and that a methodology will decrease the impact of ineffective planning, scheduling, costs control, estimating and dealing with moving targets.
This gives another spin at reasons why projects fail, other angles to be considered would be environment, requirements setting, testing, communication, the different main management topics as risks, issues, scope, quality, etc…
Another way to think about decreasing failure rate could be as in my corporation to tend to being in the Top Quartile for project delivery i.e. aspire to achieve 90% of projects delivered on time, 90% of projects delivered within budget (as set by Investment Proposal) and 90% of key business objectives met (as defined in project charter and assessed in Post Implementation Review). To achieve this criteria have been set e.g. governance, skill pool management, PMI based delivery framework and continuous improvements and this also includes end-to-end processes and measurements to allow management of product lifecycles, career paths, maturity benchmark.
In conclusion, the article only shows few aspects of better deliver business objectives and values. There is no panacea but it is necessary to have an holistic approach to this issue else only part of the solution will be developed and therefore the PM practice will not gain much credit in the business. There are multiple available methodologies, maturity models available out there but it is necessary to apply them by adapting their philosophy to the business and not the other way around, the PM practice is an enabler and not a dictator in a company.

References
KERZNER, H., 2009. Project Management: A Systems Approach to Planning, Scheduling, and Controlling. USA: John Wiley and Sons, p 63.
SEELY, M.A., DUONG, Q.P., 2001. The Dynamic Baseline Model for Project Management. Project Management Journal, Vol. 32 No.2, pp. 25-36.